Heart rate dynamics for cognitive load estimation in a driving simulation task

Abstract

Cognitive load (CL) is one of the leading factors moderating states and performance among drivers. Heavily increased CL may contribute to the development of mental stress. Averaged heart rate (HR) and heart rate variability (HRV) indices are shown to reflect CL levels in different tasks. The aim of this large-scale study was to explore how accurately HR and HRV metrics can differentiate between varying CL conditions during driving. Participants (N = 892, 44% female, from 18 to 79 years old) performed simulated driving in highway and urban scenarios. The n-back task was used as a mental distraction to further increase CL. The results have shown that increased CL was accompanied by higher HR, lower HRV, as measured by RMSSD, and higher HR complexity, as measured by permutation entropy. HR displayed the highest accuracy in discriminating between short windows (30 s) of different CL conditions, particularly highway versus urban driving and mental distraction during highway driving. We found gender and age effects on discriminative accuracy of HR and HRV metrics which were related to subjective ratings of CL. These results illustrate that HR and HRV indices provide a valid source for applications in the field of CL monitoring and mental stress detection.